New Technologies, Data Analysis and Forecasting

Managing director of research at the Society of Actuaries (SOA), R. Dale Hall, provides research insights on technology and new practices. The SOA develops and sponsors a variety of research projects and studies involving data analysis, developments in new technologies, modeling approaches, forecasting and other innovations in the insurance industry.

Can you tell us about the recent machine learning research report on insurance claims?

Hall: We wanted to better understand machine learning methods for insurance applications, so this research helps fill in gaps in how generalized linear models are used for modeling claims. This research covers a lot of ground, providing both a literature review and survey analysis on machine learning methods. This report gets more technical, exploring model formulas and parameters of the different methods. It’s helpful to see these ideas—from the approaches with random forests to regression trees—on paper. This type of research aims to expand our understanding in new areas and approaches, and I encourage you to learn more about it.

What developments are you seeing with autonomous vehicles?

Hall: We recently gathered several industry experts and actuaries to discuss automated vehicle systems and the market. We found it important to provide our members and the industry with a snapshot of automated vehicles and the implications for our profession and the overall insurance industry. Our report focuses on the potential rollout and adoption of automated systems. The researchers bucketed the automated vehicle systems into three types:

1. Vehicles with safety feature technologies
2. Self-driving systems
3. Completely driverless systems

Several factors are considered, from policy implications to public acceptance. These “trigger points” will each affect how swiftly automated systems are rolled out and adopted. It’s too soon to know all of the specifics, which is why this report helps fill in some of the gray areas and offers a variety of trigger points we should be looking for that can influence the speed of adoption. It was extremely helpful having a project review group spanning across the actuarial profession, insurance industry and government transportation agencies. Look for more information and updates as we continue to expand our understanding of the implications for the industry and the role actuaries can serve in this space.

Can you share updates on some of the recent technology and forecasting projects?

Hall: The SOA released two China-specific research projects, both of which connect with technology and data analysis projections. I’m glad to see these projects released, as they cover China-specific issues, from population estimates to technology developments affecting the insurance markets there.

For instance, one study examined the communication technology of WeChat and online marketing for insurance products in China. For those of you not familiar with WeChat, it is a rapidly growing social media application in China, with more than 300 million subscribers. It’s quickly become the go-to place to market insurance products online. Our research digs into the details of how life and health insurance companies are marketing to consumers in China through WeChat. We also found there’s increased potential through WeChat Pay options and reaching customers through customer service via the app. This is important information to know for the Chinese market, and it also may have implications to consider for the rest of the world, as insurance distribution and customer engagement grow online.

Another recent project in China forecasted the country’s population structure in the coming decades. Essentially, we looked at projections of both the mortality and fertility rates of the Chinese population. I find this analysis interesting, as it compares China’s mortality patterns to other developed countries with higher life expectancy levels and greater data quality. By 2065, a large part of China’s population will be 70 to 80 years old. There are major implications for its social security system, including a possible deficit by 2030. It’s a fascinating read, and I encourage you to check out both projects.

What other projects are in development?

Hall: I’m especially proud of the new research coming out later this year from the SOA Strategic Research Program on Actuarial Innovation and Technology. It’s one of the SOA’s five strategic research programs. Look to see some very interesting research coming up on the evolution of technology and what it means for our profession. We’re developing several research projects and studies involving technology and population trends. Stay tuned for updates as we explore how new technologies impact current and future actuarial work.

Copyright © 2019 by the Society of Actuaries, Chicago, Illinois.